Rethinking the Homework–Physical Activity Trade-Off: Cross-National Evidence from 15 High- Performing and Large Economies

Technical Summary: Rethinking the Homework–Physical Activity Trade-Off

Problem Statement
There is a widespread assumption in education and public health discourse that increased homework loads directly displace adolescent physical activity, particularly in high-performing education systems. However, cross-national evidence regarding this "time-displacement" hypothesis remains mixed. Existing literature often evaluates homework in isolation, failing to account for the interplay between academic achievement, socioeconomic status (SES), gender, and school-level context. Furthermore, many studies assume linear relationships, potentially overlooking the possibility that moderate homework loads are compatible with active routines while only extreme loads become restrictive. This study addresses the need to disentangle the effects of homework volume from broader academic achievement pressures and contextual factors to determine if homework is the primary driver of physical inactivity or if other structural factors are more significant.

Methodology
The study utilized anonymized student-level data from the Programme for International Student Assessment (PISA) 2022, covering 15 high-performing and large economies: six high-achieving Asian economies (Singapore, Macao, Chinese Taipei, Japan, Korea, Hong Kong), six high-achieving Western economies (Estonia, Canada, Ireland, Switzerland, Australia, Finland), and three large economies (USA, Germany, UK).

• Outcome Variable: Student-reported physical exercise time (item EXERPRAC), standardized within each economy to allow for cross-system comparison.
• Predictors:
  • Student-level: Academic achievement (mean of math, reading, and science scores), homework time (transformed from ordinal categories to approximate continuous hours), sex, and socioeconomic status (ESCS index).
  • School-level: Group means of achievement, homework, and ESCS.
• Analytical Approach: Economy-specific two-level linear multilevel models were estimated.
  • Variables were group-mean centered at Level 1 to isolate within-school student effects from between-school compositional effects.
  • Models included linear and quadratic terms for homework to test for nonlinearity (threshold effects).
  • Interaction terms were included to test for moderation by sex and achievement.
  • Student sampling weights were applied throughout.
  • Analyses were conducted separately for each economy rather than pooled, to preserve context-specific nuances in school organization and cultural norms.

Key Results
The analysis revealed four consistent patterns across the majority of systems, challenging the universality of the homework-displacement claim:

1. Achievement, Gender, and SES as Primary Drivers:

   • Academic Achievement: Higher academic achievement was consistently associated with less physical exercise across all 15 economies.
   • Gender: Girls reported significantly less exercise than boys in all systems.
   • SES: Students from lower SES backgrounds reported less exercise than their higher-SES peers.
   • These three factors showed stable, negative associations with exercise time, often with larger effect sizes than homework.

2. Heterogeneous and Weak Homework Effects:

   • Homework effects were generally negligible in seven economies and modestly positive in several others.
   • Only in Hong Kong was a small negative association found between homework and exercise.
   • In Chinese Taipei, Korea, Australia, and the UK, students reporting more homework actually reported more exercise.

3. Nonlinear Threshold Patterns:

   • In Singapore, Canada, and the USA, nonlinear (quadratic) models identified specific thresholds. Exercise declined only when homework levels exceeded the school-peer average by a significant margin (approximately 0.66 to 1.10 standard deviations, corresponding to roughly 2.5 to 3.2 hours per day).
   • Below these thresholds, additional homework was not associated with reduced exercise; in some cases, it was positively associated.

4. School Contextual Effects:

   • In Singapore, Japan, and Korea, attending a school with a higher average homework load was associated with slightly less exercise, even after controlling for individual homework time. This suggests a "school climate" effect where collective norms may suppress activity.
   • Contextual effects for school-level achievement and SES were inconsistent across economies, indicating that school composition impacts vary by system.

Key Contributions

• Refutation of Universal Displacement: The study provides robust cross-national evidence that the claim "homework displaces exercise" is not a universal truth. The relationship is highly context-dependent and often negligible or positive at moderate levels.
• Identification of Achievement Stratification: The paper highlights that academic achievement itself is a stronger and more consistent correlate of reduced physical activity than homework volume. This suggests that the "performance ecology" (including private tutoring, self-study, and competitive pressure) drives time reallocation more than formal homework assignments alone.
• Threshold Modeling: By identifying specific thresholds (approx. 2.5–3.2 hours/day) where negative effects emerge, the study moves beyond linear assumptions to suggest that policy should target extreme workload rather than uniform reduction.
• Multilevel Distinction: The use of group-mean centering successfully separated individual student behaviors from school-level institutional effects, revealing that school climates (e.g., in East Asian systems) can independently influence activity levels.

Significance and Claims
The authors claim that their findings necessitate a shift in policy focus from blanket homework reduction to equity-oriented, context-sensitive strategies.

• Policy Implications: Since homework reduction alone is an insufficient lever, schools should prioritize protecting movement time through scheduled physical education, active breaks, and structured after-school programs.
• Targeted Interventions: Policies must specifically address the compounded disadvantages faced by girls and lower-SES students, whose activity gaps persist regardless of homework loads.
• Systemic View: The study argues that the decline in adolescent physical activity is partly attributable to academic stratification mechanisms (tracking, admissions pressure) rather than homework per se. Therefore, educational system design (e.g., timing of high-stakes exams) must be considered alongside school-level activity mandates.
• Modesty: The authors acknowledge that their cross-sectional design precludes causal inference and that self-reported data may be subject to bias. They do not claim to have solved the issue of adolescent inactivity but rather to have clarified the relative importance of homework versus other structural and demographic factors.